After suture of proximal and distal nerve stumps into opposite ends of a fibrin-filled silicone tube, a tissue cable forms through which axons regenerate. We have continued our studies of nerve cable formation in rats, and analyzed the cellular events occurring at 17 days by electron microscopy (EM). This time was selected to determine the cellular composition of the growing tissue of each nerve end before they met in the middle of the chamber. Fibroblasts and Schwann cells formed the migratory tips of each nerve stump. Axons followed the Schwann cells in the proximal nerve whereas blood vessels grew equally well from both nerve ends. Schwann cells in the distal nerve were arranged into columns, rather than random masses of cells, and they were surrounded by basement membrane, despite the absence of axons. This columnar organization mimics the situation of in situ degenerating nerve, where Schwann cell columns guide regenerating axons. Because the endothelial cells of regenerated blood vessels in mature cables do not form tight intercellular junctions that function as an endoneurial permeability barrier, we investigated whether vascular permeability changes in in situ vessels (i.e., non-regenerated) after nerve injury. in one group of rats, the peroneal nerve was crushed so that axonal regeneration could occur through the distal nerve while, in other animals, the nerve was cut and the distal portion left permanently denervated. The vasculature of these distal nerves were permeable to an intravenously (IV) injected tracer (horseradish peroxidase, HRP) at 2 weeks after either type of injury. However, at 3 months, the distal nerve, after crush-injury, was impermeable whereas denervated nerve remained permeable to HRP. Although EM demonstrated that HRP passed between the endothelial cells, the reason for this junctional permeability was not apparent. Our studies of in situ nerve further demonstrated that normal nerve is not absolutely impermeable to vascular tracers since HRP was detected in some endoneurial fibroblasts. This activity was not endogenous since it was not seen in non-HRP injected animals. We have also begun a series of experiments in the pigeon to determine why olfactory axons regenerate back into the central nervous system whereas other sensory axons do not. it is interesting that the pigeon has a blood-brain barrier but not a blood-nerve barrier. After an IV injection of HRP, we found the tracer in the endoneurium of nerve but not the neuropil of the brain.